#shanenaz

Last year my pals and I cycled in The Galilee, Northern Israel, to raise money for Nazareth Hospital Paediatric Department. We raised over £50,000 but we could use more! Nazareth is the largest Arab town in Israel; the people are lovely, and the kids are awesome. Nazareth also treats kids in the West Bank of Palestine who have very limited access to healthcare. They need your help! Go to my sponsorship page to find out more and see what you can do! Maybe even join us in 2017..?
http://justgiving.com/shanenaz

13 November 2011

I seem to have incurred some wrath...

If there is one thing that gets right up my nasal passages, it's the encroachment of fluffy ideologies into the realms of science in general and medicine in particular. It has been my view for quite some time (and as I get older, this is concretifying rapidly, if that's even a word) that there's a group of no-doubt well-meaning but scientifically illiterate punters who are trying to develop a framework of talking about human behaviour, at all levels from the individual to the societal, based on whims and arcane concepts that have no useful mapping to anything scientific.

This is the world of pseudophilosophy (or maybe "Philosophy as She is Writ), where concepts are "isms" and everything belongs to a rigid ontological hierarchy. It's the world of CP Snow's Two Cultures, except that the scientists are the baddies, and to get anywhere you need to be able to drone at length on "post-structuralism" and "dialectics" and the like.

Anyway, within the constraints of the 140-character limit of Twitter, I became embroiled in a discussion of the value of concepts like "structure" and "agency" in addressing inequalities in health. Now I have nothing against terminlogical shortcuts per se, but the reason for my being brought into this discussion was not because I care how "social scientists" (more on that term later) view health inequalities, but because of a previous discussion with someone over what exactly "agency" was. The skinny on this is that my correspondent was trying to make the claim that "agency" implied "free will" (whatever that is). So what's this "agency" malarkey?

Some "social scientists" (commenters on public affairs, I suppose) suggest that there are two big factors in How Stuff Works in society. One is Structure, which is effectively the environment - the bricks and mortar, effectively. The other is Agency, which is what people do. Now you don't have to be a genius to figure out that these are not rigidly separate categories, because much of what is Structure is set up by the Agency of our punters (members of the ape species homo sapiens), and that Agency is influenced by the Structure, not just at the macro level, but at the very specific micro level that applies to discrete individuals. And since the Structure also has to contain the genetic background of your population, the baseline disease patterns, neural connections, personality traits, biochemical disturbances, outside influences and so on, it becomes difficult to make a good case for distinguishing Structure from Agency anyway.

Well, my ennui at this sort of cobblers spilt over to something of a diatwibe (that's a diatribe on Twitter) against "social science". Because this is where the problem really lies. "Social Science" is a spectacularly ill-defined entity; basically it acts as a rag-bag of several different disciplines, including history, archaeology, geography, that don't fit easily into what's commonly thought of as "Science". This is a shame, because the very name "Social Science" seems to lend an air of respectability to the rags within the bag, some of which deserve that respectability, and some of which manifestly do not.

It also results in the sad situation where people can call themselves "Social Scientists", and attain some sort of protection of the herd, while being able to twaddle on with nonsensical and jargon-laden pablum that has no prospect of informing policy, furthering research or leading to understanding.

I suggested (testily) that the entire literature base of social science could be destroyed, and we would be no worse off. I guess I hoped that some of the social scientists soi disant would leap to the defence of particular disciplines within the rag-bag, and indeed my original correspondent (our tussles aside) was the only one in the little Twitter group to even make a stab at disentangling the mess.

So we're stuck with what is actually the key problem. People whose understanding of science is purely based on a sociological paradigm proposing sociological models and sociological approaches to issues that deserve much more precise and evidence-based approaches. And when we are looking for evidence, rather than the opinions of "thinkers", we are into the realm of science. Proper science - not "qualitative research", not Google-bombs. We need approaches that can be tested so that conclusions - proper conclusions - can be drawn and built upon. That is something that many "social scientists" are spectacularly ill-equipped to do. Our world is complex; there are multiple feedbacks and dependencies. If we insist on adopting the top-down philosophies that percolate out of "social science" thinking, we are going to make expensive mistakes that we have no way of knowing that we're making.

Am I hammering the humanities here? No. I had the real pleasure of getting one meagre "social science" qualification, namely a Certificate in Egyptology from the University of Manchester, a couple of years back. Yes, I am told now that Egyptology is one of the social sciences. Yet within that, I found researchers and lecturers who were well versed in science and hugely sceptical of the fluffy mindset that comes from "sociology". Here were people planning proper experiments, adopting a highly scientific and rigorous approach, appropriately reticent to being too dogmatic about their conclusions, ready to be flexible and to think outside the box. Lumping Egyptology into "social science" seems such a shame.

Am I being too derogatory? Does "Social Science" proper actually exist, and does it have a role to play? Is there a real peer and post-peer review process ongoing? Is anyone leaving out the trash? No doubt your comments will help correct any of my misapprehensions...

24 comments:

  1. This comment has been removed by the author.

    ReplyDelete
  2. First, scientific illiteracy. I asked scientists to define what they do and this is what I got:

    1) Formulate theory, do experiment, show hypothesis wrong then use data from last experiment (and all the ones before) to formulate new theory
    2) We observe, we hypothesise, we test to break… We start from where we are and work outwards… But scientists do not work TO norms & values; we work WITH norms & values, and break them too… in proper science you can be *wrong*.

    If this is what scientists do, fine. There is no debate about that. Science is about explaining and predicting phenomena and about producing testable and falsifiable hypotheses. Is that correct? Theories in the social science are more evaluative and often rooted in implicit moral assumptions. Nobody is claiming objectivity. It is about interpretation and being reflexive about that interpretation.

    Second, human behaviour. The continuum of views on this, from a social science perspective, are rooted in the concepts of social order and action. Society is ordered in a top-down fashion or is the product of ongoing interactions between individuals and groups. Action is motivated by ideals, values, morals, tradition, emotional states, genetics, etc. (nonrational) or by strategic or calculated attempts to maximise rewards by minimising cost (rational). Are you saying that the only way to think about this continuum is through action being nonrational and determined exclusively by genetics? In reality, and in research, if you asked someone why they stop at a red traffic light do you think they'd say it was because of their genes? (By the way, I am all for saying it is a complex mix of all of these!)

    Structure and agency, which link to order and action discussed above, is the capacity of an individual to make a choice. In the example of the red traffic light you may argue that there really is no choice as to whether to stop (or not). However, each person will give you an answer as to why they stop. Indeed, Giddens doesn't like the duality either (which is fine). You're suggesting a unified theory of stopping at traffic lights based on a genetic code base?

    "Social Science" is a spectacularly ill-defined entity; basically it acts as a rag-bag of several different disciplines, including history, archaeology, geography, that don't fit easily into what's commonly thought of as "Science".

    Are you referring to a "unity of science" and method here? Calling it a "rag-bag" is a bit offensive but I wouldn't disagree that social science includes many disciplines that have fluid boundaries. It is an umbrella term for disciplines concerned with society. I think the rest of your post is a bit more opinionated as many social science research informs policy, furthers research and leads to a better understanding of society.

    Could you define (or prescribe) an "evidence-based approach"? I mean, social science in the main is empirical and based on observations. There are many social scientists who are commentators who don't get their hands dirty. Totally agree if that's the point you're making! I think you are misguided about the worth of qualitative research but you'll have to elaborate further on why this is the case. Qualitative research tends to look at what's really going on out there rather than being confined to the laboratory or a computer simulation. Although, agreed, both can work together and it's unhelpful to frame it as a duality.

    In terms of the notion of testing I can't see how you would "test" whether something like the National Programme for IT would be a success based on genetics? Maybe you can. Is that something you could experiment on and model in the laboratory?

    "Our world is complex", totally agree! Multiple feedbacks, totally agree. Dependencies, totally agree. No idea where you get social science as being something top-down because, as I've said, many will take both an inductive and deductive approach.

    ReplyDelete
  3. This comment has been removed by the author.

    ReplyDelete
  4. Thanks Shane. This is the commentary on structure and agency in health inequalities http://ije.oxfordjournals.org/content/39/2/378.full?keytype=ref&ijkey=9etNku0OADskgCQ
    The entire point of the article is that structure and agency are not in opposition.
    I'd like to understand more about what you think are the causes of health inequalities and what we doctors can do, if anything, about them.

    ReplyDelete
  5. Thanks folks; it's a lot more straightforward to debate these in longer text formats. Structure and Agency are not in opposition because there is no formal distinction between them. More on health inequalities later...

    ReplyDelete
  6. Right - now to deal with Mark Hawker. Mark, THANK YOU - you're one of very few of the whole pile of punters who splashed in on this little experiment to actually be man enough to put up a reasonable defence. Yes, I think you are way off-beam on some issues, but you deserve credit for giving this a shot, and being the only one who got the point. Which is perhaps understandable because you'll be aware of the history here, which most of the others aren't.

    Anyway, structure and agency. My view is that these are unhelpful and obfuscatory terms, and in order to move things forward we need to look at how they relate to other things. You might choose to relate them to nurture and nature; or environment and mentality; or even "society" and "the individual". These would be possibilities. And, in fairness, the link that AnneMarie posted originally (dealing with health inequalities for some reason; we can return to that) did make a stab at defining these.

    So if we accept that "structure" is the milieu in which the individual ("agency") finds himself or herself, is it insightful to say that we need to think of both; is it not bloody obvious that this is the case? If we want to change society, i.e. modify the decisions of individuals, we need to address the factors that influence those decisions.

    If people are to choose whether to smoke that fag or eat that burger, we can look at the availability of the noxious substance, its affordability, the education that people have, the social acceptability (which ties into the education) and a host of other factors.

    Now these do not allow us to straight away impose changes that will improve the situation we may find ourselves in. What we need to do is an experiment. We can fart about with our Grounded Theory etc, but what we really need to do is design robust experiments to trial our interventions. Then we will be able to get an idea as to whether they work, or whether they don't, and what the likely costs are going to be when we scale it up to a population level.

    All this falls within the remit of standard good old-fashioned science, and what disappoints me most is the lack of proper quantitative validation of qualitative pilots before large-scale roll-outs.

    Yes, I realise that I'm treading on territory in this response that was promised to AnneMarie as a new blog post (some day, with examples, really), but I think society has more to gain from a rigorous application of solid scientific methodologies than the sorts of analyses that tend to get attached to the "sociology" bandwagon.

    You do make some good points (and some terrible ones! ;), and I'll come back to these in the future.
    Cheers, and thanks for the tussle :-)

    ReplyDelete
  7. "If people are to choose whether to smoke that fag or eat that burger, we can look at the availability of the noxious substance, its affordability, the education that people have, the social acceptability (which ties into the education) and a host of other factors."

    How do you operationalise 'availability', 'affordability', 'education' and 'social acceptability' in your rigorous quantitative methodology? Without recourse to concepts akin to those you're dismissing?

    Your sort of simplistic. mechanistic thinking is as dangerous as the wooliest of those you're slagging off.

    I think your comment about 'a rigid ontological hierarchy' is wrong (and odd), but your view seems to exhibit a failure to acknowledge alternative epistemological approaches. There's a proverb about that (everything looking like a nail).

    ReplyDelete
  8. This comment has been removed by the author.

    ReplyDelete
  9. Your tweet said that this comment would "try to explain why I've been bashing "social science" lately". Well, it doesn't but I'm glad to read it anyway.

    Before going further I'd like to check if you're aware of the MRC guidance on developing complex interventions. Are you?

    ReplyDelete
  10. Bryan, it is not that difficult to "operationalise" these things; yes, it carries complexities, but these are well within the capacities of proper science. As for alternative epistemological approaches, this is precisely what I'm talking about. Such approaches are hypothesis generators, not hypothesis testers. Sometimes we are happy enough to go along with what looks like a good hypothesis, knowing that actually garnering proper data is too difficult, but if we elevate the epistemological reliability of qualitative approaches to that of proper science at our considerable peril (I would suggest).

    Anne Marie, I do explain it - para 2 and onwards.

    ReplyDelete
  11. This comment has been removed by the author.

    ReplyDelete
  12. "My view is that these are unhelpful and obfuscatory terms, and in order to move things forward we need to look at how they relate to other things."
    Yes. They can indeed be thought of as "society" and "the individual" if that's something we can put a box around. Is there a boundary between where "the individual" ends and "society" begins?

    "If we want to change society, i.e. modify the decisions of individuals, we need to address the factors that influence those decisions."
    Are you presuming the *only* way to modify the decisions of individuals is via changing society? What if the individuals modified their behaviours based on the close social relations around them? Indeed, Christakis and Fowler have demonstrated this in large-scale studies in both obesity and smoking cessation. And, to support your thesis that we need to relate this to "other things" then being deterministic about society is, in your own words, unhelpful.

    "If people are to choose whether to smoke that fag or eat that burger, we can look at the availability of the noxious substance, its affordability, the education that people have, the social acceptability (which ties into the education) and a host of other factors."
    See comment above. That presumes that those concepts can be both operationalised as data but that they relate in a causal way to smoking and eating burgers. If you are 100% that they have a cause/effect then you can continue with that line of thinking. You'd suggest a multiple linear regression to identify these cause/effect relationships? How large would your data set be? And, I'm assuming your concepts have been elicited via some form of social scientific inquiry through interview and observation?

    "What we need to do is an experiment."
    The experiment would be to limit availability in one town, say, and see if smoking and eating burgers reduced? That does assume people won't cross borders to buy them elsewhere. You're assuming people will be in direct compliance with your experiment? Do people in the real world work like computer simulations? If so, then good idea.

    "All this falls within the remit of standard good old-fashioned science, and what disappoints me most is the lack of proper quantitative validation of qualitative pilots before large-scale roll-outs."
    I think you're attributing the role of qualitative research as being something it's not. Even saying quantitative research would be "right" is incorrect. Though, validation and rigour is a key measure for qualitative research in the main. So too is the operationalisation of concepts in quantitative research. For example, if we get people to "rate" their happiness and they say it's a five then what does that *mean*?

    "... I think society has more to gain from a rigorous application of solid scientific methodologies than the sorts of analyses that tend to get attached to the "sociology" bandwagon."
    Indeed, there could be a "coming crisis of empirical sociology" that you can read about here: http://soc.sagepub.com/content/41/5/885.abstract.

    So, in short, I don't disagree about how science could contribute to these issues. Assuming that there are cause/effect relationships that can be isolated in society is very ambitious and generalisation could be very problematic. Validation of qualitative research is indeed very important, as is validation in quantitative research.

    ReplyDelete
  13. "Bryan, it is not that difficult to "operationalise" these things; yes, it carries complexities, but these are well within the capacities of proper science."
    How would you go about quantifying education, in your eyes? What process would you go through and would it be one data item or many? The only way I can conceive of by using "proper science" would be to identify brain activity that corresponded somehow to the concept of "education". Then again, what if you had amazing brain activity but chose not to go to university? Would you be "educated" then?

    Explain how you'd make it a quantity and that'll go some way to uncovering some more of your thinking.

    ReplyDelete
  14. It is often unclear whether you are talking about ‘social sciences’ in science (and medicine) or social sciences of science (and medicine). These both, and the charges you level at them, seem to be confused with the social sciences more generally. Most of which has very little to do with medicine or the sciences. You also seem to confuse philosophy with social science, or more accurately, social theory. Certainly social theory often looks like philosophy and one might to suggest that there is a graduation between purely empirical (positivist) social science, theoretically complex social science, social theory and philosophy, although it should be clear that this graduation is not merely a matter of degree (like a colour spectrum physically measured) but also of type (like a colour spectrum phenomenologically perceived).

    I am unclear what you might mean by the phrase “rigid ontological hierarchy” but I will say that the targets you appear to have in mind, at least at the more extreme end, would tend to oppose a rigid ontology in their theories. Indeed it would seem that it is realist science that requires a rigid ontology, which it then arranges hierarchically. You later suggest the social sciences rigidly impose a top down philosophy on its objects of study. Perhaps this is what you are referring to? Of course ontology would wrongly characterize this perspective, perhaps claiming the social sciences had a rigid metaphysics might be more accurate and still related to ontological claims. I expect, however, that a rigid epistemology, no doubt informed by certain metaphysical (including ontological perspective, is more likely to ban a accurate description of what I imagine you are trying to convey. This of course would be a largely misguided characterization of what goes on in most of the social sciences and more like what is going in on formalist philosophies of the natural sciences.

    Structure and agency are basic concepts in some social sciences. (Since you are unfamiliar with them one might liken your dismissal of the social sciences to the misguided charges you level at sociological understandings of science, but ho hum). We could might, metaphorically, liken structure and agency to nature and nurture in evolutionary biology, as you seem to do, but most social science projects tend to ‘bracket off’ concerns about the nature of the body, to treat it ‘as read’ if you like. So structure is not generally taken to include genes and organisms. Rather it is laws, social norms, language, institutions (both macro ‘political systems’, ‘healthcare systems’, median ‘schools’ ‘hospitals’ and micro ‘the family’, ‘the classroom’. Of course some social sciences are fundamentally concerned with ‘the body’ but here it is more to do with the social construction of the body, gender being the obvious example, and its effects of society and yes, science. Again the biological is largely ‘bracketed off’ as it is mostly irrelevant.

    Agency is easily read as free will, or merely will, or ‘an individual’. Very few, if any, would place it in opposition to social structure. Indeed one can consider the vast majority of social theory to be concerned with reconciling structure and agency and much of empirical social sciences to be exploring how social structure finds expression in agency. One could point to the way in which the social structures you traverse are, through the exercise of your agency, finding expression in this blog post. I would certainly say the same about me in this response. When you say there is no formal distinction between structure and agency I assume this is what you are talking about. In which case your language is misleading (at best). There is perhaps no practical or empirical distinction between structure and agency just as there is no practical or empirical distinction between nature and nurture: everything always contains both. Nevertheless there is clearly a formal or theoretical distinction.

    ReplyDelete
  15. Social science is a spectacularly ill defined entity (I take it, BTW, that your target is the interpretive and qualitative social sciences. I assume you are fine with statistics, epidemiology and other such quantitative projects). But then so is science or, to be more accurate, the natural sciences. Is it not just a rag bag of several different disciplines? Physics, chemistry, biology, psychology, earth sciences (geology, geophysics, meteorology, soil science), astronomy, engineering, electrical engineering, medicine, neurology etc. You’d have to be spectacularly ill informed to think some necessary and sufficient condition characterizes them all. Rather there is a family resemblance between them. As there is with the social sciences. Of course one can classify the sciences in a number of ways. The human sciences being my favorite. But we might also think of the behavioral sciences, formal sciences, life science, applied science. etc.

    Re. Egyptology. Having had no experience of Egyptology I am ill equipped to characterize its scientific nature. I would, however be surprised in there was no social scientific aspect to the discipline. Perhaps the closest comparison I can make is with anthropology, or at least anthropology as it is understood in America. There it is constituted by (socio-)cultural anthropology (what we would take anthropology to be in the UK), archaeology, linguistic anthropology and biological anthropology. Clearly there is a great range of both scientific and social scientific endeavors going on in most, if not all, of these strands. BTW I am not sure why you brought the humanities up in that paragraph, unless Egyptology has a humanities aspect. I would guess it probably does but that doesn’t seem to be what you were suggesting. I would also guess, therefore, that Egyptology was an inherently interdisciplinary topic rather than a single disciple per se. But one could say the same about many things, even scientific things, medicine , for example… (Indeed I once interviewed a very experienced medic who suggested they increasingly considered it t0 be a branch of the humanities. And no, before you speculate, not a GP.)

    In addition, and in an amusing parallel to those scientists who trash philosophy whilst failing to recognize the philosophical nature of their own arguments, some of your comments seem to suggest you think the social sciences a self supporting social phenomena. A practice made internally coherent by the activities of its members. This would of course be a sociological argument and require both theoretical and empirical support. The many studies done in the sociology of science and medicine will help you design such a study…

    But in answer to the question is ‘Is social science a ‘science’?’ the answer is always it depends on your definition of science. Personally I am not overly motivated to make a massive stand on the issue. I like, pace Winch, the idea of social studies rather than social sciences (Of course this conflicts with those on twitter who sought to derogate other such ‘subject studies’ as if this also implied some claim to being ‘sciences’). But equally I am sometimes ill-inclined to abandon the term ‘science’ to those who appear to think it can be fully circumscribes and its borders patrolled. Perhaps all we can conclude from this is that the term science has multiple uses in multiple contexts, much like most other words outside of formal, technical contexts (and even then…). Of course one is then left with the task of characterizing groups of sciences in more or less convenient ways, in more or less empirically and theoretically defined manners, and with reference to ones purpose in doing so. But, of course this is about where we came in. We are discussing the social sciences, after all.

    Finally, are you being too derogatory, yes. I certainly thought so repeatedly as I followed the twitter exchange. Indeed it was hard to decide if you were just trolling or if you actually thought you were engaging in constructive dialogue and argument.

    ReplyDelete
  16. Mark, thanks again for engaging. Those are valid points, and you are getting the hang of this. The secret indeed is implementing an intervention and measuring its effect. To do this we actually don't need to quantify (say) education per se - merely rank it (as a first-order attempt). Indeed, we don't even necessarily need to do that - try strategy A in one group and B in another. Or even use historical controls (with appropriate caveats - correlation & causation and all that).

    Also, I'm putting quotes around "society" deliberately - this includes family relationships & small-scale factors that may be intensely local & variable across a particular population.

    So there is no need to know what a (say) pain rating of 5 "means", other than that it's more than 4 and less than 6 to that individual; these will be variable, and we are quite prepared for our data to be messy.

    ReplyDelete
  17. "To do this we actually don't need to quantify (say) education per se - merely rank it (as a first-order attempt)."
    How would you go about ranking it? How would you scientifically ensure its validity? How would you scientifically arrive at Strategy A and Strategy B? Are you assuming that the groups would never contaminate and that people are immobile? How would you decide upon who to assign Strategy A and Strategy B to, scientifically? In short, how are you going to ensure your scientific method is scientific?

    "... this includes family relationships & small-scale factors that may be intensely local & variable across a particular population."
    Who says that's the case? Has science provided us with that definition or are you referring to a social scientific understanding of what constitutes "family relationship" and "local"? What if we are dealing with orphans? What if we are dealing with nomads and gypsies?

    "So there is no need to know what a (say) pain rating of 5 "means", other than that it's more than 4 and less than 6 to that individual; these will be variable, and we are quite prepared for our data to be messy."
    So you'd be happy for them to tell you a pain rating or would you want to monitor them 24/7 and adjust your model every nanosecond? What if they tell you they feel fine but the pain rating is really high? Will you override what they tell you? Indeed, they may relate their feeling to others around them or on other factors that are biological or social.

    OK, so far you're happy with a non-scientific method of immobile individuals based on social scientific concepts of "family relationship" and "local" and would monitor every vital sign of every individual over their entire lifetime?

    ReplyDelete
  18. Nathan, thanks for your posts above.

    ReplyDelete
  19. I think Shane is pointing out there's a huge difference between properly conducted scientific studies on well-defined problems, such as those the sort of epidemiological studies that showed the link between smoking and cancer, and the highly speculative and abstract theories of society put forward by many social scientists. The former are characterised by clearly defined problems, robust methodologies and plausible causation mechanisms, which are amenable to further scientific investigation.

    In contrast, a lot of social science theories use highly abstract and ill-defined terminology, which is essentially meaningless unless you buy into a whole raft of verbal constructs piled upon one another in something of an impenetrable wall of multi-syllabic terms. Very often these terms are disputed, ill-defined, contradictory or stratified into any number of mutually incoherent theories. Indeed it's very difficult to find much common ground in many cases. When the very grammar of a subject cannot be well defined, it rather points to a lot of people indulging pet theories.

    I should add that even in the area of epidemiology, the conclusions drawn are often represented as firmer than are merited by the evidence. No doubt this is encouraged by policy makers wishing to distill complex multi-variant problems into a simple one-dimensional message. For instance, the recommendations on alcohol units and various dietary prescriptions that presents a false sense of certainty. (Actual alcohol and otgher dietary intake recommendations vary markedly from country to country).

    Now this is not to deny that the various branches of the humanities (of which Social Sciences are one) is not of value, but to recognise the very distinct limits on the conclusions that can be reached. For instance, different social scientists come to very different conclusions over (say) the social impact of the sex industry. It doesn't actually take much reading to reveal that these are very often just manifestations of deeply held pre-set views and the use of selective evidence to promote these views. Indeed the humanities are often exercises in advocacy, not analysis. That's fair enough, but it has to be recognised as such. To my mind natural sciences are about trying to remove as much subjectivity as possible. Easy enough when dealing with (say) physics, chemistry or biology, but pretty near impossible in complex social problems.

    Too often those in "science studies" and other such parasitic subjects will claim that this attempt at scientific objectivity by the natural sciences is false, and they have no more claim to being nearer to such an ideal than any other cultural artefact.


    This latter viewpoint was dealt with quite effectively by Paul R. Gross and Norman Levitt in "Higher Superstition: The Academic Left and Its Quarrels with Science

    ReplyDelete
  20. "I think Shane is pointing out there's a huge difference between properly conducted scientific studies on well-defined problems, such as those the sort of epidemiological studies that showed the link between smoking and cancer, and the highly speculative and abstract theories of society put forward by many social scientists."
    I don't think anyone would disagree with that. However, I'd disagree on the scale in which is being implied. Indeed, "many" is a strong assertion.

    "In contrast, a lot of social science theories use highly abstract and ill-defined terminology, which is essentially meaningless unless you buy into a whole raft of verbal constructs piled upon one another in something of an impenetrable wall of multi-syllabic terms."
    Yep.

    "For instance, different social scientists come to very different conclusions over (say) the social impact of the sex industry."
    Yep. But what would science have to say about it? Could it say anything at all about morality?

    "Easy enough when dealing with (say) physics, chemistry or biology, but pretty near impossible in complex social problems."
    No denying that, either.

    "Too often those in "science studies" and other such parasitic subjects will claim that this attempt at scientific objectivity by the natural sciences is false, and they have no more claim to being nearer to such an ideal than any other cultural artefact."
    I agree with you, as with Shane, up until here. Who are "those" and "they"? Where is your scientific evidence of this or are you just basing it on your own experience? In which case, you are being a social scientist with N=1.

    ReplyDelete
  21. This comment has been removed by the author.

    ReplyDelete
  22. If your claims are based on making "postmodernism" and "social science" and "science studies" equal then I'm afraid that book is a major straw man. As pointed out by nathan this has already been a major point of confusion.

    ReplyDelete
  23. Let me get this straight, so far we have "unearthed" that people doing social science may put forward "abstract theories", sometimes use long words, can be subjective and come up with different interpretations.

    If that's all that's being put forward then I could have read that in a first-year text book. I mean, really?

    ReplyDelete
  24. Hi folks, and thanks, especially Mark and Nathan for your considered responses. Yes, I admit I was trolling a bit yesterday (quite a bit, actually), and I also admit I was putting forward a viewpoint that could be regarded as a tad extreme. I know a few people were well pissed off with what I had to say; some of them *just* left it at the pissed off level, but what I was hoping was that at least someone might be able to step up.

    Nathan, those are good points well made (as are Mark's) - I wouldn't include statistics as a "social science", but I get where you're going. One point I picked up on that I think deserves a bit of examination is the concept of things being multi-disciplinary. Yes, I've represented a somewhat extreme view - even a caricature to an extent - above, but in my defence I would say that many "hard" scientists recognise elements of this, and at the very least it functions as a barrier to proper multidisciplinary working.

    I suppose another aspect is that very many scientists do have interests, hobbies, projects ongoing in fields that would perhaps be regarded as "social science" (if not "sociology of science" or "science studies"), and there may be fertile ground there for some exploration. I do feel, however, that we need to be working on methods to put some social science onto a better scientific footing - not "sciency" spurious precision (such as informs much of the nutritional guff that gets onto cereal packets), but appropriately constructed frameworks that allow and encourage well designed experiments to be carried out.

    Anyway, thanks again folks - this has been interesting, and despite my Victor Meldrew impersonation, your comments are welcomed and valued.

    -S

    ReplyDelete